Keyboard Test

Test every browser-visible key, see what is held or already verified, measure simultaneous rollover, flag suspicious rapid double-fires, inspect repeat timing, and export a local session report without sending your keystrokes anywhere.

Wave182 · browser-local keyboard diagnostic

Keyboard Test — coverage, rollover, chatter & repeat

Press physical keys and watch exactly what the browser receives. Verify every visible key, test simultaneous rollover, inspect suspicious rapid double-fires, measure hold/repeat behavior, and export a local session report.

private · no keystroke upload

Paused

Start capture, then press the physical keyboard. Escape releases capture so browser shortcuts work again.

0/0Keys tested in selected layout

0Held now

0Max simultaneous

0Max non-modifier keys

0Potential issues

0:00Session duration

Last browser event

—event.key

—event.code

—location

—repeat

—modifiers

—event / hold

Physical code vs typed character: event.code follows the physical key position; event.key follows your active keyboard layout and modifiers.

Rollover combo check

Common combination

0/0 received now

Hold the listed keys physically at the same time. If one does not appear, the keyboard matrix, firmware, connection, OS, or browser interception may be limiting the combination.

Hold & key-repeat

—First repeat delay

—Avg repeat interval

0Logged events

Repeat delay/interval reflects OS/browser key-repeat delivery, not USB polling rate or switch scan rate. Hold duration is browser-visible keydown→keyup time.

Potential issue suspects

A rapid second keydown can be intentional fast tapping. A long hold can be intentional too. Treat these as suspects to retest, not automatic hardware-failure verdicts.

Latest key events

Session time	Key / code	Event	Modifiers

Optional local session summaries

Saved summaries contain counts and layout only — not the raw event log or characters you typed.

Browser boundary

This tool sees only KeyboardEvent data delivered to this tab. Reserved shortcuts, Fn/media layers, firmware remaps and OS accessibility features may intercept or transform keys before the page receives them. It cannot measure USB polling rate, firmware scan rate, electrical switch bounce, or click-to-photon input latency.

Press Start capture. Escape releases keyboard capture.

How to diagnose a keyboard without overclaiming

Dead-key check is visual coverage

Choose the closest layout and press every physical key. Untested keys are keys this browser session has not received — they are not automatically “dead.” Retest in another application before blaming the switch.

Rollover and ghosting are combination problems

A keyboard can pass every key individually and still drop one member of a multi-key chord. Hold the exact gaming or shortcut combinations you care about and compare your fingers with Held now and Max simultaneous.

Chatter is a suspect, not a verdict

Mechanical chatter can create unusually fast duplicate presses, but deliberate fast tapping can look similar. Use the configurable threshold to find suspects, then press the same key slowly several times to confirm the pattern.

OS repeat timing is measurable; hardware polling is not

Hold one key to see when browser repeat begins and its average interval. Those values reflect the operating system and browser event stream. They are not a direct 125/500/1000 Hz USB polling measurement.

Full-size through compact coverage

Full ANSI, TKL, 75%, 65% and 60% views change the visual checklist so you can focus on keys your board is expected to have. Compact views are coverage profiles, not a promise that every manufacturer uses the same physical key geometry.

Privacy & export

Key events stay in this browser page. CSV/TXT exports are created locally. Saving a session is optional and stores only counts/layout; it does not save the raw key history or typed text.